The Risk of Cholesteatoma in Individuals With First-degree Relatives Surgically Treated for the Disease

Key Points Question Is cholesteatoma in a first-degree relative associated with an increased risk of the disease? Findings In this nationwide case-control study of first-time cholesteatoma surgeries including 10 618 cases and 21 235 controls in Sweden, the risk of cholesteatoma surgery for individuals with a first-degree relative treated for the disease was almost 4 times increased, but few cases were exposed. Meaning This study suggests that a strong hereditary component in cholesteatoma disease exists, but this only explains a limited number of cholesteatoma cases.


C
holesteatoma in the middle ear is an uncommon disease accompanied by risks for severe complications if not treated.2][3] There is a male dominance in having had cholesteatoma surgery. 4 cholesteatoma is typically defined as an acquired retraction pocket in the tympanic membrane that loses its ability to self-clean and thus starts to accumulate keratin debris. 5,6With time, the cholesteatoma grows and affects surrounding bone and other nerve and soft tissues, which can lead to hearing loss and disturbances of taste, balance, and facial nerve function.Infection of the cholesteatoma is common and seems to increase the rate of bone resorption. 7Cholesteatoma, if not treated, can lead to sinus thrombosis, meningitis, and intracranial abscess.
Several other types of cholesteatoma exists, including (1) acquired nonretraction pocket cholesteatoma emerging from a perforated tympanic membrane, (2) congenital cholesteatoma under an intact membrane, and (3) postsurgical cholesteatoma divided in recurrent or residual cholesteatoma. 6The treatment of the disease is surgical, but occasionally individuals will not undergo surgery, mostly due to medical reasons.
In the literature, cholesteatoma is in general not presented as a hereditary disease, but in clinical practice, histories of familial cases are accumulating.The results of a search in the literature are sparse, and in a 2018 review article by Jennings et al, 8 only 35 articles were found on the subject, ranging from case reports of siblings to genetic testing.However, in 2009, Prinsley 9 reported on family clustering of cholesteatoma in 12 families in the United Kingdom, the largest cohort yet reported in the literature.In 2019, a preliminary communication from the same research group 10 showed the results of a genetic study of one of these families with the identification of 2 genes of interest.
To our knowledge, no previous population-based studies have examined heritability of cholesteatoma in a large scale.In Sweden, medical care is tax-funded and available for all citizens.Moreover, medical procedures are recorded in nationwide health registers that can be linked to other population registers using the unique personal identity number that is assigned to all residents in Sweden. 11This infrastructure enables the possibility for nationwide studies of rare diseases.In this study, the aim was to investigate the risk of cholesteatoma in individuals with a first-degree relative surgically treated for the same disease.

Study Design
This is a nationwide population-based case-control study nested within the Swedish total population between 1987 and 2018.The study was conducted by record-linkage of several Swedish national health data and population registers.The study has been approved by The Swedish Ethical Review Authority (2019-05190, 2020-000245, 2021-05727-02), and in-formed consent was waived due to the large sample size and the use of register data.The study follows the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.
The rating of the quality of evidence was 3 due to the casecontrol study design.However, the study includes national, high quality register data covering a whole nation during a 30-year time period.

Cases and Controls
Cases of cholesteatoma were identified from the Swedish National Patient Register. 12This register includes discharge diagnoses, as well as information on procedures performed, from all inpatient and specialized outpatient care.The coverage of inpatient care has been nationwide since 1987, and outpatient visits to specialized care (physicians) have been included since 2001. 12The register has in general a high sensitivity in regard to surgical interventions. 12All individuals with a diagnosis of cholesteatoma treated with cholesteatoma surgery between 1987 and 2018 (see eTable in the Supplement for diagnostic codes according to International Classification of Diseases, Ninth Revision [ICD-9] and International Statistical Classification of Diseases and Related Health Problems, Tenth Revision [ICD-10] and Swedish procedural codes) were defined as cases of cholesteatoma and included at the time of first-registered cholesteatoma surgery during the study period.Cases were further divided according to surgical approach of the cholesteatoma removal into atticus or mastoid, and other locations, as indicated in the registered procedural codes (eTable in the Supplement).
Through incidence density sampling, 2 controls per case were randomly selected from the general population using the Total Population Register 13 and were matched to cases by age, sex, and municipality of residence at the date of surgery of the case (hereinafter, index date).The control selection was conducted by Statistics Sweden.

Exposure
The exposure was defined as having a first-degree relative surgically treated for cholesteatoma.For cases and controls, all first-degree relatives (ie, biological parents, full and half siblings, and children) were identified using the Multigeneration Register 14 (Figure).This national register enables linkage between children and parents for persons born in 1932

Key Points
Question Is cholesteatoma in a first-degree relative associated with an increased risk of the disease?Findings In this nationwide case-control study of first-time cholesteatoma surgeries including 10 618 cases and 21 235 controls in Sweden, the risk of cholesteatoma surgery for individuals with a first-degree relative treated for the disease was almost 4 times increased, but few cases were exposed.
Meaning This study suggests that a strong hereditary component in cholesteatoma disease exists, but this only explains a limited number of cholesteatoma cases.and later. 14In the next step, all first-degree relatives that were registered in Sweden any time between 1987 and 2018 were linked to the National Patient Register to assess if they had been treated with cholesteatoma surgery (using the same definition as for the cases).In the main analysis, cases and controls were considered exposed if at least 1 first-degree relative had been treated with cholesteatoma surgery before or after the index date.Cases and controls were considered unexposed if at least 1 first-degree relative was identified but had no record of cholesteatoma in the National Patient Register.Additionally, the exposure was defined by relationship (ie, having a parent [mother and father separately], sibling, and child) with cholesteatoma.
To elucidate the role of increased awareness of the condition (eg, if a close family member had been treated for cholesteatoma it might be more likely that the condition is diagnosed and treated also in the index person), the exposure was defined as having a partner that had been treated with cholesteatoma surgery (before or after the index date).Partners to cases and controls were identified from the Total Population Register in the year preceding the index date, or within the same year.Partners were defined as marital partners for the full study period, and additionally as cohabitants with common children (from 1990 and onwards) or cohabitants without common children (from 2011 and onwards).

Statistical Analysis
Baseline characteristics were compared between cases and controls using Fisher exact test, and a P value ≤.05 was considered statistically significant.The association between having a first-degree relative with cholesteatoma and the risk of cholesteatoma surgery in the index persons were estimated by calculating odds ratios (ORs) and 95% CIs using conditional logistic regression.These analyses are inherently adjusted for the matching factors age, sex, and region of residence.As the controls are drawn at the time of surgery (ie, by incidence density sampling), the ORs can be interpreted as incidence rate ratios. 15e main analysis assessed the risk of cholesteatoma surgery by having any first-degree relative with cholesteatoma (yes or no), and analyses were also performed separately by relationship.Since familial cholesteatoma may arise earlier in life, the analyses were stratified by age at index date (<20 years old or ≥20 years old).The diagnosis of cholesteatoma could only be assessed from 1987 and onwards, among both index persons and relatives, and to assess the potential effect of this left truncation of the data, the analyses were stratified by index period (1987-1999 or 2000-2018).Finally, the analyses were stratified by cholesteatoma surgery extension (atticus or mastoid vs other).The estimates of the stratified analyses were compared using the method described by Altman and Bland. 16o assess the potential effect of increased health awareness, the risk of cholesteatoma was compared between individuals with and without a partner with cholesteatoma.
All analyses were restricted to cases and controls who had at least 1 first-degree relative identified, that were registered in Sweden any time between 1987 and 2018, in the category of interest (eg, mother, father, child, sibling).Data were received in April 2022, and analyses were conducted between April and September 2022.All analyses were conducted in IBM SPSS Statistics for Windows, version 28.0.0.0 (IBM) or SAS statistical software, version 9.4 (SAS Institute Inc).

Results
Between 1987 and 2018, 10 618 individuals with a first-time cholesteatoma surgery were identified in the Swedish National Patient Register.Men accounted for 6302 (59.4%) individuals, and mean (SD) age at surgery was 35.6 (21.5) years (34.3 [21.2] for men and 37.4 [21.7] for women, Table 1).Most cases of cholesteatoma were found between the ages of 8 and 18 years old (eFigure 1 in the Supplement).In addition to the cases, 21 235 matched controls were included.The mortality in cases and controls were similar; 1472 cases (13.9%) and 2791 controls (13.1%)died during the study period.In total, 146 202 first-degree relatives to cases and controls were identified (that were registered in Sweden at any time between 1987 and 2018), with a mean value of 4.6 relatives per index person (for both cases and controls).For 10 133 cases (95.4%) and 20 412 controls (96.1%), at least 1 firstdegree relative were identified (Table 1).
Among the 10 105 cases with at least 1 control included in the main analysis (Figure ), 227 [2.2%] had at least 1 firstdegree relative treated for cholesteatoma, while the corresponding numbers for controls were 118 of 19 553 control patients (0.6%, Table 2).The risk of cholesteatoma surgery among individuals with a first-degree relative treated for the disease, compared with individuals with no affected firstdegree relatives was estimated at an OR of 3.9 (95% CI, 3.1-4.8)(Table 2, eFigure 2 in the Supplement).The association differed slightly by sex of the index person, and the OR was higher among men (OR, 4.2; 95% CI, 3.1-5.6)than women (OR, 3.4; 95% CI, 2.4-4.9).The increased risk of cholesteatoma surgery among individuals with affected first-degree relatives were seen for all relationships, with effect sizes spanning between 2.9 (95% CI, 1.9-4.6)for having a child treated for cholesteatoma to 4.4 (95% CI, 3.2-6.2) for having a sibling treated for the disease.
Table 3 presents the analysis stratified by age, type of surgery, and time period of surgery.The OR was higher for individuals with cholesteatoma surgery performed under the age of 20 years (OR, 5.2; 95% CI, 3.5-7.5)as compared with individuals aged 20 years or older (OR, 3.2; 95% CI, 2.4-4.3), and for cases where surgery involved the atticus and/or mastoid areas (OR, 4.9; 95% CI, 3.6-6.6)compared with other loca-  Abbreviation: OR, odds ratio.a The analyzed population includes only cases and (their corresponding) controls where a relative could be identified, therefore the total number differs compared with Table 1.

Discussion
This population-based case-control study shows an almost 4-fold-increased risk for cholesteatoma in individuals having at least 1 first-degree relative surgically treated for the disease, compared with individuals with no affected relatives.The association was particularly strong in young individuals and those whose surgery involved atticus and/or mastoid areas.No association was found when comparing the likelihood of having a partner treated for cholesteatoma.This strengthens the hypothesis of a hereditary disease rather than an association driven by increased health awareness.
This nationwide, register-based, case-control study in Sweden included a large number of patients and investigated the risk of family clustering of cholesteatoma during a 30-year long period.Utilizing the Swedish health and population registers, all individuals surgically treated for a cholesteatoma were included, independent on surgical unit performing the operation, which minimizes the risk of selection into the study.Moreover, the tax-funded Swedish health care system offers an equal medical coverage for all citizens minimizing a socioeconomically induced selection bias.All first-degree relatives registered in Sweden during this period were identified, and their disease history, assessed through health data registers; thus, there is no misclassification due to self-reporting of exposure.Some case reports of familiar clustering of cholesteatoma are found in the literature 9,17,18 indicating a possibility of a hereditary pattern.In 2014, Djurhuus et al 19 found a 2-fold risk for cholesteatoma surgery in siblings to individuals with cleft palate but no difference in the groups with lip or combined lip and palate cleft, compared with a random sample of the Danish population.Such distinction cannot be made in the current study's data because no information of malformations was available.
Having a first-degree relative with cholesteatoma increased the risk of a first cholesteatoma surgery particularly in the younger ages (<20 years at surgery).This could imply that inheritance is of higher importance for childhood cholesteatomas.1][22][23] How-  Abbreviation: OR, odds ratio.a The analyzed population includes only cases and (their corresponding) controls where a partner could be identified, therefore the total number differs compared with Table 1.
ever, the register data used in the current study could not distinguish between congenital and acquired cholesteatoma due to the lack of unique ICD codes, making this impossible to investigate without additional information from medical records.Yet, a recent meta-analysis 24 regarding congenital cholesteatoma stated that the mean age at surgery was 4.9 years, and in the current study, the most common age at cholesteatoma surgery was 10 years (eFigure 1 in the Supplement).The number of surgeries in patients younger than 7 years in Sweden is very low, accounting for 6.8% of all surgeries (eFigure 1 in the Supplement).The incidence of congenital cholesteatoma is calculated to be between approximately 4% to 24% of all cholesteatomas and the incidence has been shown to be increasing over the last years in South Korea. 25,26This said, the influence of a possible inheritance in congenital cholesteatomas affecting the current study's results cannot be excluded.
The results of the current study indicated slightly more pronounced associations among men compared with women, although the differences were not statistically significant.Male patients have a higher incidence of both otitis media and cholesteatoma in the population, [1][2][3]27 but the results in this study also point toward a higher susceptibility to inheritance. Sevral studies have shown a strong genetic component for otitis media in general, 28,29 and a high percentage of individuals surgically treated for cholesteatoma has a history of otitis media.Due to limitations in register data, the frequency of childhood otitis media was not possible to investigate.The heredity seen in this study could therefore be explained, completely or partly, by the heredity for otitis media and thus have similar origin.
In 2 large cohort studies, the incidence of cholesteatoma was increased in children treated with ventilation tube insertion for otitis media, 30,31 and several studies have shown a higher incidence in individuals with craniofacial anomalies, such as cleft palate, Turner syndrome, and Down syndrome. 19,32,33Except for a higher incidence of cholesteatoma in siblings to individuals with cleft palate, 19 no studies have been performed for the other subgroups in order to investigate their part in the pattern of inheritance.Further studies in this area are needed.
Having a first-degree relative with cholesteatoma was accompanied with a particularly increased risk for cholesteatoma surgery involving the atticus or mastoid region.This may indicate a different mechanism for cholesteatomas originating in the mesotympanic area.More studies combining register and clinical data are needed to further investigate a difference in cholesteatoma type and inheritance pattern.

Limitations
Despite a strong relative association between having a firstdegree relative with cholesteatoma and the risk of the dis-ease, the number of exposed cases was low, 227 of 10 105 cases (2.2%).This implies that the association only explains a limited number of cholesteatoma occurrences.Interpreting the absolute prevalence of family history of cholesteatoma should be done with caution due to a lack of information regarding cholesteatoma surgeries before 1987 in the current study.This means that cholesteatoma surgery among relatives occurring earlier than 1987 are lacking, and thus, some exposed individuals are misclassified as nonexposed giving an underestimation of the absolute prevalence of family history of cholesteatoma.However, this has outcomes for both cases and controls in the same way, and because this is a rare exposure, the potential effect on the relative estimates are at the most only a minor dilution.This left truncation of the data also implied that some of the cases in the early time period might be included in the study at their second instead of at their primary surgery.
Cases were identified using both diagnostic and procedure codes.This strengthens the sensitivity of the case ascertainment compared with using only diagnostic codes.However, this also means that cases not surgically treated or misdiagnosed are not included, which might lead to a minor loss of precision of the current study's estimates.Using procedure codes and registry data also leads to a possibility of an overdiagnosis if surgeons have used a cholesteatoma diagnosis when performing surgery for reconstruction purposes secondary to a cholesteatoma surgery.However, that would only overestimate the number of cases during the early time period and to some extent identify the false negative (ie, cases treated in the years prior to 1987).However, results from analyses stratified by time periods were largely similar, indicating only a minor potential effect from these potential biases.

Conclusions
This case-control study shows an almost 4-fold risk estimate for cholesteatoma in individuals with a first-degree relative surgically treated for the disease.This association was not explained by an enhanced awareness of the condition within families, as no association was found with having a partner with the outcome.Although the relative association was strong, family history was nevertheless quite rare and can therefore only explain a limited number of all cholesteatoma cases.However, these families could be an important source for information regarding the genetic background for cholesteatoma disease.The use of nationwide register data with high coverage and completeness strengthens the result.Future studies including second-degree relatives and information from medical records could further illuminate the association of heredity.

Figure .
Figure.Flowchart of Included Study Population and Analytical Sample

Table 1 .
Descriptive Characteristics of Patients Treated With Surgery for Cholesteatoma During 1987-2018 in Sweden and Matched Controls From the General Population

Table 2 .
Odds Ratios of Cholesteatoma Associated With Having at Least 1 First-degree Relative Surgically Treated for Cholesteatoma, in Total and by Type of Relative a

Table 3 .
Odds Ratios of Cholesteatoma Associated With Having at Least 1 First-degree Relative Surgically Treated for Cholesteatoma, Stratified by Time Period, Age, and Type of Surgery P values are estimated with the Altman and Bland 16 method. a

Table 4 .
Odds Ratios of Cholesteatoma Associated With Having a Partner Surgically Treated for Cholesteatoma Among Cases and Controls a